225 
Polycycly. 
most highly developed form of polycyclic dictyostely. Thus they 
come naturally at the end of the series of cases we have been 
discussing, though it is quite certain that they have no direct phylo¬ 
genetic connexion with any of the Ferns we have already dealt with. 
The polycyclic dictyostely of the modern Marattiaceas is shared by 
the well-known Carboniferous and Permian tree-ferns, the Psaronieae, 
so that we are obliged to believe that the factors giving rise to these 
highly complex vascular systems came into play at a very early 
period, long before they had become operative in any of the great 
Leptosporangiate series, perhaps before the Leptosporangiate type 
appeared at all. It is all the more interesting to find that these now 
isolated forms show, in connexion with the appearance of these 
complex characters, phenomena essentially the same as those we 
have been considering in the Leptosporangiate Ferns. 
The well-known tree-fern stems from the Carboniferous and 
Permian, which, when they are preserved so as to show anatomical 
structure, are placed in the genus Psarouius, are almost certainly of 
Marattiaceous affinity. Both the evidence of association with 
Marattiaceous sporangia, and also many anatomical details, point in 
this direction. In habit, however, the Psaronieae were lofty tree- 
ferns, even taller than the modern Cyatheaceae, and in the general 
structure of their vascular systems, they have something in common 
with this family, though scarcely enough to justify the conclusion of 
Stenzel, and of their most recent investigator, Rudolph (05), that 
they have a real affinity with the modern tree-ferns. 
Like the living Angiopteris, the Psaronieae are highly polycyclic, 
from five to seven distinct vascular rings occurring in the stems of 
some forms; but, unlike those of the Marattiaceae, the outermost 
vascular cylinder is not perforated, so that the meristeles are for 
the most part broad and band-shaped (Fig. 70), indicating, like 
those of the Cyatheaceae, a slighter departure from the primitive 
solenostelic type. In the form of the leaf-trace also, Psarouius is 
relatively primitive, its cross-section being C-shaped with incurved 
ends, rather like that ot a Dicksonia, and indeed the structure of 
the complicated vascular system of the stem, and its relation to 
the leaf-traces, show a close resemblance to the solenostelic 
Saccolonia {Dicksonia Lindeni Hk). 
The species' of Psarouius are divided by Zeiller (’90) into three 
’ i.e., distinct types of structure. It may very well be, as Zeiller 
points out, that in some cases different “species” may 
represent different stages of development of one form, and 
on the other hand, more than one species in the ordinary 
sense may be comprised in a single anatomical “ species.” 
